When so many incidents share one particular outcome — armed civilians stopping attackers — it seems difficult to dismiss the pattern as accidental.
Data ignored by the FBI show the importance of defensive gun use in bringing down active shooters.
T wo days after last week’s attack on an Oregon elementary school, the Associated Press opened a news story with a familiar claim: “Few mass shootings in the U.S. end like the deadly attack on an Oregon elementary school: with an armed civilian opening fire on the gunman.” NPR also noted “how rarely an armed civilian ends an attack like this.”
Yet since January 2025, the Crime Prevention Research Center (CPRC) has documented 16 defensive gun uses in which police credited armed civilians with stopping what otherwise might have become mass public shootings, as in the attack in Oregon. During that same period, five mass public shootings occurred in the U.S. and Puerto Rico.
The AP bases its claim on data from the Advanced Law Enforcement Rapid Response Training (ALERRT) Center, which the FBI uses for its active-shooting reports. But the story ignores detailed challenges to those data that have appeared in such places as the Washington Post, Washington Times, and RealClearInvestigations.
The FBI defines an active shooter as someone actively killing or attempting to kill people in a populated, public area. An incident can qualify even if a gunman fires at just one person and misses. The category includes mass public shootings, but the FBI generally excludes shootings connected to other crimes, such as robberies or fights over drug turf.
I encountered problems with these data firsthand. Until January 2021, I served as senior adviser for research and statistics in the Justice Department’s Office of Justice Programs, where my responsibilities included evaluating the FBI’s active-shooter reports. I discovered that the reports omitted or misclassified numerous cases in which armed civilians stopped attacks.
The problem persists. From 2014 through 2025, the FBI identified 408 active-shooter incidents but credited armed citizens with stopping only 16 — just 3.9 percent.
The CPRC, however, has identified numerous additional cases and maintains an updated list. Our research tells a dramatically different story. We identified 615 active-shooter incidents from 2014 through 2025 and found that armed citizens stopped 225 of them. We excluded another 35 cases in which law-abiding armed citizens stopped attackers before they could fire a shot.
By our count, armed citizens stopped 36.6 percent of the active-shooter incidents identified over those twelve years, nearly ten times the FBI’s 3.9 percent. Over the most recent five years, that figure rose to 42.5 percent.
The percentage rises further when we consider only places where law-abiding citizens may legally carry firearms. After all, you can’t expect armed civilians to stop many attacks in places that prohibit them from carrying. Outside gun-free zones, we estimate that armed civilians stopped 56.3 percent of active-shooter attacks over the past five years.
In 2023, Washington Post fact-checker Glenn Kessler asked the FBI and ALERRT about the missing cases. The FBI declined to respond, while ALERRT challenged the inclusion of two of the hundreds of incidents on our list.
The first involved a man who opened fire in a dental office in Colonial Heights, Tenn., in 2019, killing his wife. ALERRT argued that the incident did not qualify because the husband had targeted his wife at work during a domestic dispute. The second involved a man who opened fire in a Portland, Ore., strip club in 2014 after being kicked out. ALERRT said that incident “sounded like a retaliation murder.”
Yet ALERRT includes domestic and retaliatory shootings when attackers also endanger others, as both of these cases clearly do. Our research identifies 14 cases in ALERRT’s data that otherwise mirror the Tennessee attack: An attacker went to a significant other’s workplace and opened fire. The only difference? Armed civilians did not stop any of the incidents ALERRT included.
We found the same inconsistency with the Oregon case. ALERRT included other attacks in which someone returned with a gun after getting kicked out of a bar, but it excluded a similar attack that an armed civilian stopped.
In 2024, journalist John Stossel asked the FBI about the completeness of its data. The bureau acknowledged that its data were “not intended to explore all active shooting incidents but rather to provide a baseline understanding.” Yet the FBI omits that qualification from its reports and press releases.
Anyone can miss a case. But when so many incidents share one particular outcome — armed civilians stopping attackers — it seems difficult to dismiss the pattern as accidental. Even more troubling, the FBI compounds the problem by leaving errors uncorrected after they are brought to its attention.
Readers need not take the CPRC’s word for it. We list each case and link to the underlying sources, so anyone can examine the evidence and compare our findings with the FBI’s reports.